Vol. 117 No. 2 February 2014

Classification and characteristics of pterygoid process fracture associated with maxillary transverse fracture Jingang An, MD,a Fanfan Dai, MD,b Zhipeng Sun, MD,c and Yi Zhang, MDd School and Hospital of Stomatology, Peking University, Beijing, China

Objective. This study aimed to classify pterygoid process fractures associated with maxillary transverse fractures. Study Design. Pterygoid process fractures in 100 patients with maxillary transverse fractures were observed 2- and 3-dimensionally using image processing software. Fracture line course and height and sphenoid sinus involvement were recorded. Results. Pterygoid process fractures were classified as follows: class I, vertical (simple separation between medial and lateral plates); or class II, transverse (3 subcategories according to location of fracture line: II-1, within pterygoid fossa; II-2, above pterygoid fossa, not extending to sphenoid sinus floor; II-3, above pterygoid fossa, involving sphenoid sinus floor). Class I fracture was observed on 5 sides (2.7%); II-1, on 125 (66.5%); II-2, on 36 (19.1%); and II-3, on 22 (1.7%). Conclusions. Pterygoid process fractures were predominantly near the upper edge of the pterygoid fossa. Pneumatization of the pterygoid process is a risk in fractures involving the sphenoid sinus floor. (Oral Surg Oral Med Oral Pathol Oral Radiol 2014;117:243-252)

The LeFort classification of maxillary transverse frac- study, medical image processing software was used ture remains in wide use after almost a century. The to view 2- and 3-dimensional computed tomogra- common feature of the 3 types of LeFort fracture is phy (CT) images of maxillary transverse fractures that the sphenoid pterygoid process is involved1,2; from different perspectives. The characteristics of indeed, some authorities believe that the presence of a the pterygoid process fractures associated with the pterygoid process fracture almost always predicts the maxillary transverse fracture were observed and presence of a LeFort maxillary fracture.3 However, combined with clinical data to classify pterygoid because it is located deep in the posterior part of the process fracture. maxilla, certain aspects of pterygoid process fracture have not been reported in detail in the literature, MATERIALS AND METHODS owing to the limitations of the imaging techniques The participants were patients with maxillary transverse in the past. An example is the pattern of pterygoid fracture who were admitted to the Oral and Maxillo- process fracture lines associated with maxillary facial Surgery Trauma Center at the School and Hos- transverse fracture: are they also transverse, or do they pital of Stomatology, Peking University, between involve separation between the medial and lateral January 2007 and October 2011. All patients included plates or between the pterygoid process and the in the study met the following criteria: (1) maxillary maxilla? Do the different levels of LeFort fracture fracture was recorded in the medical records and correlate with different levels of pterygoid process confirmed by preoperative CT, and the surgical records fracture? In the previous literature, the characteristics showed that maxillary fracture reduction and fixation of the pterygoid process fracture caused by maxillary were performed; (2) complete clinical data were avail- LeFort I osteotomy were discussed, and some simple able, including detailed medical records, preoperative fi classi cations were made, but literature discussing the facial and occlusal photos, and preoperative CT of the classification of pterygoid process fracture associated 4-8 head; (3) the time between the injury and admittance with maxillary LeFort fractures was rare. In this was less than 2 months, no surgical fracture reduction or fixation had been performed, and the maxillary aAssociate Professor, Department of Oral and Maxillofacial Surgery, Peking University School and Hospital of Stomatology. bResident, Department of Oral and Maxillofacial Surgery, Peking University School and Hospital of Stomatology. Statement of Clinical Relevance cAttending Physician, Department of Oral Radiology, Peking Uni- versity School and Hospital of Stomatology. This article summarizes the characteristics of ptery- d Professor, Department of Oral and Maxillofacial Surgery, Peking goid process fractures associated with maxillary University School and Hospital of Stomatology. fi Received for publication Apr 13, 2012; returned for revision Nov 13, transverse fractures and provides a classi cation for 2013; accepted for publication Nov 26, 2013. pterygoid process fractures. The result can help oral Ó 2014 Published by Elsevier Inc. and maxillofacial surgeons and radiologists under- 2212-4403/$ - see front matter stand the features of pterygoid process fractures. http://dx.doi.org/10.1016/j.oooo.2013.11.500

243 ORAL AND MAXILLOFACIAL RADIOLOGY OOOO 244 An et al. February 2014

Table I. LeFort classification and characteristics of bilateral maxillary fractures (89 cases) Single-line fractures Multiline fractures Bilateral consistent Bilateral inconsistent Bilateral consistent Bilateral inconsistent I II III I þ II I þ III II þ III I þ II þ III 14 7 4 13 8 1 3 13 26 15.7% 7.9% 4.5% 14.6% 9.0% 1.1% 3.4% 14.6% 29.2%

transverse fracture had been classified according to the Table II. Relationship between maxillary fracture and LeFort classification. pterygoid process fracture (cases) A total of 100 cases met the inclusion criteria for this Unilateral pterygoid Bilateral pterygoid study (86 male, 14 female), with a mean age of 33.7 Type process fracture process fracture Total years (range, 14-59 years). The causes of the fractures Unilateral maxillary 11 0 11 were traffic accidents (75 cases), work-related injuries fracture (13), falls (9), blast injury (2), and violent injury (1). Bilateral maxillary 18889 The mean time from injury to admittance was 19.3 fracture days, with a range of 1 to 50 days. Total 12 88 100 Head CT images were reconstructed using image processing software (Mimics, version 10.01; Materialise, fractures and pterygoid process fractures is shown in Leuven, Belgium), and maxillary LeFort fractures were Table II (c2 test; P < .05). These findings indicate that identified from axial, coronal, sagittal, and 3-dimensional there is a correlation between maxillary fracture and images. Pterygoid process fractures were observed on pterygoid process fracture, unilateral maxillary fracture axial, coronal, and sagittal images. To obtain recon- being associated with unilateral pterygoid process structed 3-dimensional maxillofacial images, the man- fracture and bilateral maxillary fracture with bilateral dible, the cervical vertebrae, and the rear of the skull, pterygoid process fracture. which obstruct views of the pterygoid process from the Based on the characteristics of the pterygoid process posterior and both sides, were removed; the back and fractures observed in this sample, the following classi- sides of both pterygoid processes were then completely fication was established: class I, pterygoid process visible in 3 dimensions. The information obtained from vertical fracture (simple separation between the medial the 2-dimensional and 3-dimensional images allowed us and lateral plates); or class II, pterygoid process trans- to determine the pattern of pterygoid process fracture verse fracture, which was further divided into 3 sub- lines, fracture level, and extension to the sphenoid sinus, categories: II-1, fracture line located within the pterygoid and ultimately to classify pterygoid process fracture fossa; II-2, fracture line located above the pterygoid associated with maxillary transverse fracture. The default fossa, not extending to the floor of the sphenoid sinus; II- threshold (226-3071HU) for bony tissue in Mimics was 3, fracture line located above the pterygoid fossa and used to segment these CT images. involving the floor of the sphenoid sinus (Figures 1 to 5). Statistical analysis was conducted with statistical According to this classification, the 100 cases (188 software (SPSS, version 11.0; SPSS Inc, Chicago, IL, sides) comprised 138 sides with simple fractures (73.4%) USA) using the c2 test and the nonparametric Kruskal- and 50 sides with complex fractures of 2 or 3 types Wallis test. Statistical significance was set at P < .05. (26.6%); 48 involved 2 types, and 2 involved 3 types. Thus, there were 240 fracture sites in total (Table III). RESULTS Of the 50 complex fractures, 44 were single transverse The 100 cases had a total of 189 sides with maxillary fractures with separation between the medial and fractures, with unilateral fracture in 11 (11 sides) and lateral plates (Figure 6); 4 were transverse fractures bilateral fracture in 89 (178 sides). The characteristics both above and below the upper edge of the pterygoid of the fracture lines in the 89 bilateral maxillary frac- fossa (Figure 7); and 2 were transverse fractures both tures are shown in Table I. The 189 sides with maxil- above and below the upper edge of the pterygoid fossa lary fractures in this sample corresponded to 188 sides with separation between the medial and lateral plates. with pterygoid process fractures. The pterygoid process The distribution of pterygoid process fractures, fracture was unilateral in 12 cases (12 sides), 11 of including simple and complex fractures, is shown in which had a unilateral maxillary fracture (with 1 case of Table IV. bilateral maxillary fracture), and bilateral in 88 cases Among the 138 sides of simple pterygoid process (176 sides), with 96 left sides and 92 right sides fracture, there were 75 sides of single-line maxillary affected. The relationship between the maxillary fracture and 63 sides of multiline maxillary fracture. OOOO ORIGINAL ARTICLE Volume 117, Number 2 An et al. 245

Fig. 1. Diagram of a coronal section of the right pterygoid plate, showing the classification of pterygoid plate fractures. A, Class I, vertical fracture. B, Class II-1, transverse fracture, fracture line located within the pterygoid fossa. C, Class II-2, transverse fracture, fracture line located above the pterygoid fossa and not extending to the floor of the sphenoid sinus. D, Class II-3, transverse fracture, fracture line located above the pterygoid fossa with sphenoid sinus floor involvement. (SS, sphenoid sinus; PF, pterygoid fossa.)

Fig. 2. Pterygoid process class I fracture. Vertical fracture of the pterygoid lateral plate; no fracture in the medial plate.

Among the 50 sides of complex pterygoid process Among the 51 cases of bilateral maxillary fracture in fracture, there were 22 sides of single-line maxillary which both fractures were located at the same level, the fracture and 28 sides of multiline maxillary fracture 2 pterygoid process fracture lines were located at the (Table V)(c2 test; P > .05). These results suggest that same level in 36 cases and at different levels in 15 there is no relationship between the comminution de- cases. Among the 38 cases of bilateral maxillary frac- gree of the pterygoid process fracture and that of the ture in which the 2 fractures were located at different maxillary fracture. levels, the 2 pterygoid process fracture lines were Based on the highest fracture line on each side, class I located at the same level in 24 cases and at different fractures were identified on 5 sides (2.7%), II-1 on 125 levels in 14 cases (Table VII)(c2 test; P < .05). These sides (66.5%), II-2 on 36 sides (19.1%), and II-3 on 22 results suggest that the level of bilateral maxillary sides (1.7%). Most of the class II-1 fracture lines were fracture lines does not predict the level of the associated located near the upper edge of the pterygoid fossa; the II- bilateral pterygoid process fracture lines. 2 fracture lines were located above the pterygoid fossa Of the 100 cases in this sample, 36 had sphenoid but also close to the upper edge of the pterygoid fossa. sinus fractures. Of those, 23 had associated fractures in Together, class II-1 and II-2 accounted for 161 sides in the nasal-orbital-ethmoid region, with CT showing total (85.6%). The maxillary fractures were analyzed in a frontal, ethmoid, and sphenoid sinus fractures similar manner (Table VI). The Kruskal-Wallis test (Figure 8), and 5 (7 sides) had class II-3 pterygoid found no relationship between the height of the maxil- process fractures; the other 13 cases had sphenoid sinus lary fracture line and that of the pterygoid process frac- floor fractures, and all (15 sides) had class II-3 ptery- ture line (P > .05), which means that when the maxillary goid process fractures. The sites of the sphenoid sinus fracture is located high, the corresponding pterygoid fractures were the sinus floor in 18 cases (22 sides), the process fracture is not necessarily high, and vice versa. anterior wall in 18, the lateral wall in 8, the sinus ORAL AND MAXILLOFACIAL RADIOLOGY OOOO 246 An et al. February 2014

Fig. 3. Bilateral pterygoid process class II-1 fracture. Bilateral transverse fractures; the fracture lines are located within the pterygoid fossa, below the upper edge of the pterygoid fossa. septum in 2, and the top wall in 1. The fracture pterygoid fissure, containing the pyramidal process of extended close to the middle cranial fossa in 1 case. the palatal plate. The superior and anterior parts of the Among the 100 cases, the sphenoid sinus air cell pterygoid process are separated from the posterior of spread into the root of the pterygoid process in 29 cases the maxilla by the pterygomaxillary fissure, whereas the (54 sides); that is, there was pneumatization of the root inferior and anterior areas of the pterygoid process of the pterygoid process. Of these 29 cases, 25 were connect with the posterior part of the maxilla to form the bilateral and 4 were unilateral. Eighteen cases (22 sides) pterygomaxillary ligament at the pterygomaxillary had class II-3 pterygoid process fractures, of which 14 junction.9 Thus, the superior part of the sphenoid pter- (18 sides) had a large sphenoid sinus and pneumatiza- ygoid process connects to the base of the middle cranial tion of the corresponding pterygoid process; the other 4 fossa, whereas its anterior and inferior parts connect to (4 sides) showed no pneumatization. The correlation the posterior of the maxilla. Anatomically, the pterygoid between pneumatization of the pterygoid process and process is closely related to the maxilla, and together class II-3 fracture for all 189 sides is shown in they constitute the midfacial vertical pterygomaxillary Table VIII (c2 test; P < .05). The results indicate that buttress. Some authors suggest that, although they are pneumatization of the pterygoid process is a risk in anatomically components of the basisphenoid, the high-level pterygoid process fractures that extend to the pterygoid processes are functionally more closely allied sphenoid sinus floor. with the facial skeleton, and pterygoid process fracture may be more appropriately considered a facial injury.10 DISCUSSION Through anatomic and histologic studies, Melsen and The sphenoid pterygoid process extends inferiorly from Ousterhout11 found that at the inferior part in the the junction between the body of the sphenoid and the calcifying areas of the maxillary tuberosity, the pyra- greater sphenoid wing. Each pterygoid process com- midal process of the palatine bone and the pterygoid prises 2 plates, the medial and the lateral. The lateral process of the sphenoid bone were seen to approach plate is wide and thin, with its lateral surface facing each other through a sheet of loose fibrous tissue. But in anteriorly and exteriorly. The medial plate is long and the late juvenile and the adolescent stages, there were narrow; its lower end turns sharply and curves laterally closed complex connections along these bones, and at its inferior margins into a hook-like process called the disarticulation invariably led to bony fractures. pterygoid hamulus. The 2 plates fuse anteriorly and Unger et al.10 reported 111 cases of craniofacial superiorly, and they diverge inferiorly to form the fractures, 78 of which had fractures of the sphenoid OOOO ORIGINAL ARTICLE Volume 117, Number 2 An et al. 247

Fig. 4. Bilateral pterygoid process class II-2 fracture. The fracture lines are located above the pterygoid fossa and below the base of the skull, not extending to the floor of the sphenoid sinus.

Fig. 5. Left-sided pterygoid process class II-3 fracture. The fracture line is located above the pterygoid fossa, extending to the floor of sphenoid sinus, where it led to an effusion. bone; 31 of these (51 sides) were pterygoid process zygomaticomaxillary complex fractures. In the present fractures, 45 sides were associated with LeFort maxil- study, in addition to the maxillary transverse fractures, lary fractures, and 2 sides were associated with fractures we also found a small number of zygomatic fractures of the zygomatic bone. Unger and Unger12 described 12 caused by a lateral striking force that were associated cases of pterygoid process fracture, 9 of which were with pterygoid process fractures. These cases were not associated with LeFort maxillary fractures and 3 with included in the analyzed sample, which contained only ORAL AND MAXILLOFACIAL RADIOLOGY OOOO 248 An et al. February 2014

Table III. Distribution of types of pterygoid process fracture among 188 sides Simple fracture Complex fracture Total I II-1 II-2 II-3 I þ II-1 I þ II-2 I þ II-3 II-1 þ II-2 II-1 þ II-3 I þ II-1 þ II-3 592251633101 1 3 2 188

Fig. 6. Left-sided pterygoid process class II-1 fracture associated with a class I fracture.

Fig. 7. Right-sided pterygoid process class II-1 fracture associated with a class II-3 fracture. pterygoid process fractures associated with maxillary of this fracture is now technically guaranteed. Orhan transverse fracture. et al.14 evaluated the pterygoid hamulus morphology Because of the close relationship between the maxilla using cone beam computed tomography, and the and the sphenoid pterygoid process in the rear, maxil- knowledge about the morphology of these structures lary transverse fracture will inevitably extend to the provides valuable information in the differential diag- pterygoid process; thus, this type of fracture can be nosis of untraceable pains in the oral cavity and phar- called a maxillaepterygoid process complex fracture. ynx. Coronal CT is considered optimal for observing The characteristics of pterygoid process fractures pterygoid process fractures, including the fracture level associated with maxillary fractures are rarely reported and sphenoid sinus floor involvement; axial CT can because of their deep location. Deep fractures are also show whether the medial and lateral pterygoid plates known as occult fractures and are generally difficult to are separated and whether the pterygoid maxillary display clearly on plain films. Before the advent of CT junction has diverged; and 3-dimensional CT can fully imaging technology, tomography was used clinically to visualize the characteristics of these fractures. However, diagnose pterygoid process fractures, but unsatisfacto- the pterygoid process cannot be observed directly on rily.13 With the clinical use of CT and continuing maxillofacial 3-dimensional CT images, because the progress in imaging technology, a good understanding view is blocked by the surrounding bones. To study OOOO ORIGINAL ARTICLE Volume 117, Number 2 An et al. 249

Table IV. Distribution of types of pterygoid process Table VI. Distribution of pterygoid process fractures fracture according to the level of maxillary fracture (sides) Classification I II-1 II-2 II-3 Total Classification II-1 II-2 II-3 Total Simple fracture 5 92 25 16 138 LeFort I 37 7 4 48 Complex fracture 46 39 11 6 102 LeFort II 39 10 7 56 Total 51 131 36 22 240 LeFort III 49 19 11 79 Total 125 36 22 183

Table V. Relationship between pterygoid process fracture lines and maxillary fracture lines Table VII. Relationship between bilateral maxillary fracture level and bilateral pterygoid fracture level Maxillary Maxillary Type (sides) single-line fracture multiline fracture Total Pterygoid Pterygoid fractures at fractures at Pterygoid process, 75 63 138 Classification (cases) same level different levels Total single-line fracture Pterygoid process, 22 28 50 Bilateral maxillary fractures 36 15 51 multiline fracture at same level Total 97 91 188 Bilateral maxillary fractures 24 14 38 at different levels Total 60 29 89 pterygoid process fractures, software is used to remove the posterior and bilateral bones from images. fractures were defined as those that occurred at or near A relationship was found between maxillary trans- the base of the skull, low-level fractures were those that verse fracture and pterygoid process fracture, in that occurred at the level of the osteotomy cut, and inter- most unilateral maxillary transverse fractures were mediate fractures were those that occurred slightly accompanied by unilateral pterygoid process fractures, above the level of the osteotomy cut. Lanigan6 found whereas bilateral maxillary transverse fractures were vertical fractures of the pterygoid plates extending associated with bilateral pterygoid process fractures. In toward the skull base, which is similar to the findings in general, we could therefore infer the presence of a our class I fractures. Precious et al.8 summarized the maxillary transverse fracture from a pterygoid process characteristics of pterygoid plate fractures in 58 cases of fracture, especially in patients with nonobvious mal- maxillary LeFort I osteotomy and found that the frac- position, and pterygoid process fractures could be used tures all occurred in the lower half of the pterygoid as indirect diagnostic evidence of maxillary transverse process. fractures. If CT scans of a patient with maxillofacial We want to determine whether the characteristics of injuries clearly identify a pterygoid process fracture, pterygoid process fracture associated with maxillary a maxillary transverse fracture is likely to have transverse fractures and those of pterygoid process occurred, and maxillary dental arch malposition and fracture in LeFort I osteotomy are the same. A defini- malocclusion requiring surgery are probably present tive classification of pterygoid process fracture is (Figure 9). needed, because there is none in the literature of facial When maxillary LeFort I osteotomy is performed, the trauma. The pterygoid process is a slender bone struc- procedure of separation of the maxillary tuberosities ture comprising medial and lateral plates that are from the pterygoid plates can cause pterygoid process increasingly fused from the base up. At the base of the fracture. In the literature, the incidence of pterygoid process, the medial and lateral plates diverge, with process fracture secondary to LeFort I osteotomy the pterygoid fissure between them; in the middle part, ranged from 20% to 87%, and the fracture sites were the medial and lateral plates fuse anteriorly and diverge observed and classified. Robinson and Hendy4 carried posteriorly, with the pterygoid fossa between them. At out LeFort I osteotomy on 8 cadavers and found that its top, the pterygoid process connects with the base there were low-level fractures and high-level fractures. of the skull, and the medial and lateral plates are The low-level fractures occurred at the level of the completely fused. Through observation of the pterygoid osteotomy cut, and the high-level fractures occurred process fractures in this study, we found that vertical close to the base of the skull. Renick and Symington5 fractures led to separation of the medial and lateral also divided the pterygoid fractures into low-level plates, and the fracture line did not reach beyond the fractures, which were inferior to the LeFort I osteotomy pterygoid fossa. In most cases, pterygoid process line, and high-level fractures, which were superior to transverse fractures occurred near the upper edge of the this level. Lanigan6 and Laster et al.7 divided the pterygoid fossa, especially below the upper edge. Those pterygoid process fractures into 3 types: high-level that occur above the pterygoid fossa should be ORAL AND MAXILLOFACIAL RADIOLOGY OOOO 250 An et al. February 2014

Fig. 8. A patient with LeFort I and II maxillary fractures, maxillary sagittal fracture, and nasal-orbital ethmoid fracture. On axial and sagittal computed tomography images, the fractures are seen extending to the ethmoid and sphenoid sinus, from the anterior backward. Neither pterygoid process fracture (right II-1, left II-2) extended to the floor of the sphenoid sinus. Sphenoid sinus fractures of this type are caused by a frontal striking force.

Table VIII. Relationship between pneumatization of in the midportion in LeFort II fractures, and in the base the pterygoid process and class II-3 fracture in 188 in LeFort III fractures. Osborn16 suggested that, in sides with pterygoid process fracture LeFort I and II fractures, associated pterygoid process Category II-3 fracture Non-II-3 fracture Total fractures occur in the inferior or midportion of the pterygoid process, whereas in LeFort III fractures they Pterygoid process 18 36 54 pneumatization occur at the junction between the pterygoid process and No pterygoid process 4 130 134 the body of the sphenoid. In both studies, low-level pneumatization maxillary fractures were associated with low-level Total 22 166 188 pterygoid process fractures, and high-level maxillary fractures were associated with high-level pterygoid process fractures, which is not in accord with the considered high-level pterygoid process fractures, findings of the present study. regardless of whether they extend to the floor of the Through review of the cases in our sample, we found sphenoid sinus. Based on the characteristics of their that there was no relationship between the level of the fracture lines, we established a classification of ptery- pterygoid process fracture line and that of the maxillary goid process fractures. Hopper et al.15 suggested that fracture line. Most of the pterygoid process fractures fracture of the posterior wall of the maxillary sinus were class II-1 (66.5%) or II-2 (19.1%); this means that without pterygoid process fracture may coexist with they occurred predominantly near the upper edge of the maxillary transverse fracture; however, we did not pterygoid fossa and indicates that this area may be the observe this situation in our sample, nor any separation structurally weakest area of the pterygoid process. Of between the pterygoid process and the maxilla. the 188 sides with pterygoid process fractures, 138 LeFort I, II, and III maxillary fractures differ in (73.4%) had a simple fracture and 50 (26.6%) had fracture line height. How, then, does the height of the complex fractures of 2 or more subtypes. Simple pter- associated pterygoid process fracture lines vary? Unger ygoid vertical fractures were relatively rare, occurring and Unger12 found that pterygoid process fractures in only 5 sides. Pterygoid vertical fractures associated occurred at the junction of the middle and inferior third with transverse fractures were more common, occurring of the pterygoid process in LeFort I maxillary fractures, in 46 sides (24.5%) and indicating that a considerable OOOO ORIGINAL ARTICLE Volume 117, Number 2 An et al. 251

Fig. 9. A young female patient with facial trauma. A, A preoperative 3-dimensional computed tomography image showed bilateral pterygoid process class II-1I fractures associated with class I fractures. B, A 3-dimensional computed tomography image showed right-sided maxillary LeFort I and II fractures and left-sided maxillary LeFort III fractures. The left-sided maxillary fracture line was not clear, so the left-sided pterygoid process fracture can be used as indirect evidence for the left-sided maxillary fracture. C, A preoperative open bite deformity in the anterior teeth region suggested that bilateral maxillary transverse fracture existed. proportion of transverse fractures occur simultaneously fossa. Of the 36 cases of sphenoid sinus fracture in this with separation of the medial and lateral plates. study, only 13 (15 sides) were caused by pterygoid Through observation and statistical analysis (see process class II-3 fractures; the other 23 were complex Table V), we found that there was no relationship be- fractures of the frontal, ethmoid, and sphenoid sinuses, tween the comminution degree of the pterygoid process 5 of which (7 sides) also had pterygoid process II-3 fracture and that of the maxillary fracture. Because the fractures. This indicates that, in sphenoid sinus fractures number of maxillary fracture lines to some extent rep- associated with midface fractures, force could be con- resents the intensity of the frontal force on the maxilla, ducted directly from the anterior backward or indirectly we preliminarily speculated that the comminution de- from an inferior pterygoid process fracture. A consid- gree of the pterygoid process fracture not only is related erable portion of sphenoid sinus fractures are caused by to the intensity of the frontal force on the maxilla but a frontal striking force that extends backward through also is related to the form, volume, and strength of the the frontal and ethmoid sinuses (weak areas of the base pterygoid process. of the skull) to the sphenoid sinus, causing its fracture In this sample, 22 sides (11.7%) had class II-3 frac- and even involving the middle cranial fossa. Based on tures. This is a high-level transverse fracture extending these data, it is suggested that, compared with pterygoid to the floor of the sphenoid sinus. Simple II-3 fractures process fracture, a frontal striking force leads more associated with fractures of other parts of the middle easily to sphenoid sinus or middle cranial fossa fracture. cranial fossa were not observed. Pneumatization For the middle cranial fossa, the danger posed by a sometimes extends into the pterygoid process and frontal striking force is greater than that of an inferior plates, and there is a structural weak area at the root of pterygoid process fracture. the pterygoid process. CT scan examination can be used to identify this anatomic variation clearly.17,18 Statisti- cal analysis found that excessive gasification of the REFERENCES 1. Katzen JT, Jarrahy R, Eby JB, Mathiasen RA, Margulies DR, sphenoid sinus was a risk factor for II-3 fractures; that Shahinian HK. Craniofacial and skull base trauma. J Trauma. is, the risk of high-level fractures extending to the floor 2003;54:1026-1034. of the sphenoid sinus was higher in patients with 2. Ghobrial W, Amstutz S, Mathog RH. Fractures of the sphenoid excessive gasification within the base of the pterygoid bone. Head Neck Surg. 1986;8:447-455. process. The base of the pterygoid process and the floor 3. Rhea JT, Novelline RA. How to simplify the CT diagnosis of LeFort fractures. AJR Am J Roentgenol. 2005;184:1700-1705. of the sphenoid sinus are considered to be the struc- 4. Robinson PP, Hendy CW. Pterygoid plate fractures caused by the turally weakest parts of the middle cranial fossa floor, LeFort I osteotomy. Br J Oral Maxillofac Surg. 1986;24:198-202. and high-level pterygoid process fractures leading to 5. Renick BM, Symington JM. Postoperative computed tomography fracture of the sphenoid sinus floor could prevent an study of pterygomaxillary separation during the LeFort I osteot- injurious force from extending upward, thereby pro- omy. J Oral Maxillofac Surg. 1991;49:1061-1065. 6. Lanigan DT, Loewy J. Postoperative computed tomography scan tecting the brain. Pterygoid process fractures caused by study of the pterygomaxillary separation during the LeFort I maxillary transverse fractures therefore generally would osteotomy using a micro-oscillating saw. J Oral Maxillofac Surg. not cause substantial damage to the middle cranial 1995;53:1161-1166. ORAL AND MAXILLOFACIAL RADIOLOGY OOOO 252 An et al. February 2014

7. Laster Z, Ardekian L, Rachmiel A, Peled M. Use of the ‘shark-fin’ 15. Hopper RA, Salemy S, Sze RW. Diagnosis of midface fractures osteotome in separation of the pterygomaxillary junction in with CT: what the surgeon needs to know. Radiographics. LeFort I osteotomy: a clinical and computerized tomography 2006;26:783-793. study. Int J Oral Maxillofac Surg. 2002;31:100-103. 16. Osborn AG. Radiology of the pterygoid plates and pter- 8. Precious DS, Goodday RH, Bourget L, Skulsky FG. Pterygoid ygopalatine fossa. AJR Am J Roentgenol. 1979;132:389-394. plate fracture in LeFort I osteotomy with and without pterygoid 17. Carter LC, Pfaffenbach A, Donley M. Hyperaeration of the chisel: a computed tomography scan evaluation of 58 patients. sphenoid sinus: cause for concern? Oral Surg Oral Med Oral J Oral Maxillofac Surg. 1993;51:151-153. Pathol Oral Radiol Endod. 1999;88:506-510. 9. Chong VF, Fan YF, Tng CH. Pictorial review: radiology of the 18. Liang H, Benson BW, Frederiksen NL. Pneumatization of the sphenoid bone. Clin Radiol. 1998;53:882-893. pterygoid process of the sphenoid bone. Dentomaxillofac Radiol. 10. Unger JM, Gentry LR, Grossman JE. Sphenoid fractures: prev- 2001;30:63. alence, sites, and significance. Radiology. 1990;175:175-180. 11. Melsen B, Ousterhout DK. Anatomy and development of the pterygopalatomaxillary region, studied in relation to LeFort : osteotomies. Ann Plast Surg. 1987;19:16-28. Reprint requests 12. Unger JD, Unger GF. Fractures of the pterygoid processes accom- Yi Zhang, MD panying severe facial bone injury. Radiology. 1971;98:311-316. Department of Oral and Maxillofacial Surgery 13. Archer CR, Sundaram M. Uncommon sphenoidal fractures and Peking University School and Hospital of Stomatology their sequelae. Radiology. 1977;122:157-161. 22 Zhongguancun Nandajie 14. Orhan K, Sakul BU, Oz U, Bilecenoglu B. Evaluation of the Haidian district pterygoid hamulus morphology using cone beam computed to- Beijing 100081 mography. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. China 2011;112:e48-e55. [email protected]